BIM-based green building design method

Abstract

The application discloses a kind of green building design methods based on BIM, belong to building design technical field.For solving the problems, such as the internal thermal disturbance parameter acquisition in existing design depends on artificial, parameter completion is not unified, simulation result is not credible, design modification depends on artificial and leads to low iteration efficiency, the method is by constructing IFC format BIM model, extracts and completes geometric information and thermal performance parameter, automatically obtains internal thermal disturbance parameter using semantic matching, independently simulates and cross-verification using main and auxiliary dual energy consumption simulation engine, when exceeding, automatically decomposes energy consumption exceeding amount to component, generates structured modification instruction write back BIM model and triggers a new round of simulation.The application can be used for energy consumption simulation and automatic optimization iteration of green building design.

Description

Technical Field

[0001] This invention relates to the field of architectural design technology, and more specifically to a BIM-based green building design method. Background Technology

[0002] In BIM (Building Information Modeling)-based green building design practices, building energy consumption simulation analysis is a crucial step in evaluating whether the design scheme meets energy-saving and green performance requirements. However, the data transfer process from the BIM model to the energy consumption simulation model has long been hampered by several obstacles, affecting the iterative efficiency of the design phase. First, obtaining internal thermal disturbance parameters for buildings heavily relies on manual operation. Data such as personnel density, lighting power density, equipment power density, and corresponding hourly usage schedules required for energy consumption simulation typically require designers to manually input them into simulation software after consulting code manuals for each functional zoning of the building. Although the BIM model stores the geometric and name information of spatial objects, the existing process cannot automatically convert this information into usable internal thermal disturbance parameters, making the data preparation phase time-consuming and prone to inconsistencies. Secondly, the lack of geometric information and thermal performance parameters of the building envelope is a common problem in BIM models. During the design refinement process, the detailed dimensions of some components are not yet determined, and the thermal properties of some materials are not clearly defined in the model. Current practices typically involve simulation analysts supplementing or defaulting values ​​based on experience. However, the standards for supplementation vary from person to person, leading to differences in the simulation input conditions for the same building by different analysts, thus affecting the comparability of simulation results across different scenarios. Furthermore, the reliability of energy consumption simulation results lacks an effective verification mechanism. In current practice, the calculation results of a single simulation engine are often used directly as the basis for design decisions. When there are simplification assumptions in the model or deviations in the input parameters, the simulation results may deviate significantly from the actual operating performance of the building, and designers find it difficult to determine the degree and source of this deviation. Furthermore, when simulation results indicate that energy consumption indicators fail to meet green building targets, the design modification process still relies on manual intervention. After interpreting the simulation report, analysts need to return to the BIM modeling environment to manually adjust component parameters, re-export the model, and resubmit the simulation. This "analysis-modification-reanalysis" cycle is fragmented, lacking automated data transfer and feedback mechanisms, thus increasing the design iteration cycle. Summary of the Invention

[0003] This invention provides a BIM-based green building design method, aiming to solve a series of technical problems existing in the current BIM-based green building design process. Specifically, the main problems to be solved include: in the existing process, the internal thermal disturbance parameters of the building rely on manual input one by one; the standards for supplementing missing geometric information and thermal performance parameters of the building envelope are inconsistent; the energy consumption simulation results lack a credibility verification mechanism; and design modifications cannot be automatically feedback in a closed loop, resulting in low design iteration efficiency. Furthermore, when the thermal performance of the building envelope components is optimized to the performance limit of the product parameter library, if the building energy consumption index still does not meet the requirements for green buildings, the existing method lacks an automated alternative optimization path, and the optimization process is prone to stagnation. At the same time, in the process of reducing solar radiation heat gain by reducing the window-to-wall ratio, the existing method lacks automatic verification of indoor natural lighting performance, which may lead to contradictions between energy-saving measures and lighting compliance requirements. Furthermore, when calling the size parameters of external shading components, existing methods do not consider the cross-space shading effect of adjacent window components on the same facade, which may lead to the actual shading effect of the external shading components deviating from the design expectation. When multiple adjacent spaces on the same facade independently generate new instructions for external shading components, existing methods lack a coordination mechanism, and independent adjustments to the shading components in each space may lead to mutual shading or superimposed effects, resulting in iterative oscillations. Regarding model credibility verification, when dual-engine cross-validation fails, existing methods rely on manual investigation of the sources of differences in the input parameters of the two models, a time-consuming process requiring professional experience. During model self-calibration, adjusting the timetable parameters purely for mathematical convergence may produce distorted parameter combinations that numerically meet tolerance requirements but do not conform to the actual usage patterns of the building. Furthermore, using a fixed shape deviation tolerance threshold cannot adapt to the varying requirements of timetable flexibility for different building types, and a single threshold cannot simultaneously consider the feasibility of calibration convergence and the rationality of building function. Moreover, there are internal differences in the activity patterns of people on different floors and facing areas within the same building, and a uniform hierarchical threshold for the entire building cannot accurately match the actual usage patterns of each area.

[0004] To achieve these objectives and other advantages according to the present invention, a BIM-based green building design method is provided, comprising the following steps: Step 1: Construct the BIM model of the target building. The BIM model is an IFC (Industry Foundation Classes) format file and contains geometric information of building components, thermal performance parameters of the building envelope, building material property information, and spatial objects expressed by IfcSpace instances. Step 2: Extract geometric information and thermal performance parameters of the building envelope from the IFC format file, convert the geometric information into gbXML (Green Building XML) geometric description data, map the thermal performance parameters of the building envelope to gbXML material attribute fields, and complete any missing geometric information and thermal performance parameters. Step 3: Extract the name and description attributes of IfcSpace instances from the IFC format file, and use a pre-trained semantic matching model to determine the spatial function type corresponding to each IfcSpace; obtain the building type from the IfcProject attribute. When this attribute is empty, determine the building type based on the area ratio of each spatial function type; using the building type and spatial function type as indexes, obtain the personnel density, lighting power density, equipment power density, and corresponding schedule for each IfcSpace from the preset internal thermal disturbance parameter rule base, and write them into gbXML format data. Step 4: Import the gbXML format data after completing the geometric information, thermal performance parameters and internal thermal disturbance parameters into the main energy consumption simulation engine and the auxiliary energy consumption simulation engine based on different mathematical mechanisms, respectively, and independently execute the hourly energy consumption simulation throughout the year, and output the first energy consumption index data and the second energy consumption index data respectively. Step 5: Calculate the normalized mean bias and root mean square error coefficient of variation between the first energy consumption index data and the second energy consumption index data. When both the normalized mean bias and the root mean square error coefficient of variation are within their respective preset engineering tolerance thresholds, the first energy consumption index data is used as the final energy consumption index data; otherwise, the model self-calibration is performed until the tolerance requirements are met. Step Six: Compare the final energy consumption index data with the preset green building energy consumption threshold range. When the final energy consumption index data exceeds the green building energy consumption threshold range, decompose the excess energy consumption to each building envelope component to form a key component list. Generate component-level structured modification instructions based on the preset high-performance component parameter library and write them into the BIM model through the application programming interface of the BIM modeling software. This will automatically trigger a new round of gbXML file export and building energy consumption simulation calculation.

[0005] Preferably, in the BIM-based green building design method, when it is detected that the current thermal performance parameter of a component in the key component list has reached the upper limit of the performance of the same type of component in the preset high-performance component parameter library, and the final energy consumption index data still exceeds the green building energy consumption threshold range, the following steps are further performed: The parameters of the component are locked as the upper limit of performance. The building orientation and the facade identification of the component are extracted. The excess energy consumption difference of the component that exceeds the green building energy consumption threshold range in the energy consumption excess decomposition in step six is ​​converted into an equivalent solar radiation heat gain reduction target value. The equivalent solar radiation heat gain reduction target value is obtained by dividing the excess energy consumption difference by the preset comprehensive energy efficiency ratio of the air conditioning system. Based on the target value of solar radiation heat gain reduction, the corresponding window-to-wall ratio adjustment range or external shading component size parameters are queried from the preset passive component parameter adjustment rule library, and a window-to-wall ratio modification instruction or external shading component addition instruction containing a unique identifier of the window component and an exterior facade identifier is generated. The window-to-wall ratio modification command or the external shading component addition command is written into the BIM model through the application programming interface of the BIM modeling software, triggering a new round of building energy consumption simulation calculation.

[0006] Preferably, the BIM-based green building design method further includes a natural lighting compliance verification step before generating the window-to-wall ratio modification instruction. Extract the geometric boundary information and visible light transmittance parameters of the space to which the window component belongs from the BIM model. Based on the sky illuminance data in the typical meteorological year file, calculate the percentage of full natural lighting time on the reference working plane of the space. When the percentage of total natural daylighting time is lower than the preset threshold for green building daylighting compliance, the window-to-wall ratio modification instruction is rejected, the original window-to-wall ratio parameter is retained, and the adjustable external shading component size parameter corresponding to the window component is preferentially retrieved from the passive component parameter adjustment rule library according to the target value of solar radiation heat gain reduction. The external shading component addition instruction is regenerated and written into the BIM model.

[0007] Preferably, in the BIM-based green building design method, when retrieving and calling the adjustable external shading component size parameters corresponding to the window component from the passive component parameter adjustment rule library, a cross-space shading coordination step is also included: Extract the location coordinates and geometric dimensions of the window component in the adjacent space on the same facade as the window component from the BIM model; Project the window components of the adjacent space into the current verification space in parallel, and calculate the percentage of the overlapping area between the projected area and the window components of the current verification space. When the overlap area exceeds the preset shading threshold, the size parameters of the retrieved adjustable external shading component are reduced according to the preset reduction coefficient, and a new instruction for external shading component is generated based on the reduced size parameters.

[0008] Preferably, in the BIM-based green building design method, when multiple adjacent spaces on the same facade simultaneously generate new instructions for external shading components, the method also includes a main control step for coordinating adjacent spaces: Extract the window area parameters of all external shading components to be added on the facade from the BIM model, and determine the space with the largest window area as the dominant space of the facade. Prioritize the execution of new external shading components for the dominant space, and use the size parameters of the external shading components for the dominant space as the reference shading parameters for that facade. When generating new instructions for external shading components in non-dominant spaces, the reference shading parameters are substituted as known boundary conditions into the cross-space shading coordination step for reduction calculation, and the reduced size parameters are used to generate new instructions for external shading components in non-dominant spaces.

[0009] Preferably, in the BIM-based green building design method, step five, model self-calibration, includes: Extract the time parameters used to define the occupancy rate, lighting utilization rate, and equipment utilization rate from the main energy consumption simulation engine and the auxiliary energy consumption simulation engine, respectively. Within the percentage range of their respective timetable parameters, the timetable parameters in the two simulation engines are adjusted synchronously with a preset step size to generate multiple sets of timetable parameter combinations and re-execute the annual hourly energy consumption simulation until the calculated normalized mean error and root mean square error coefficient of variation are both within their respective preset engineering tolerance threshold ranges. The final set of timetable parameters that meets the tolerance requirements, along with their corresponding first energy consumption index data, will be used as the final reliable simulation input and output results.

[0010] Preferably, the BIM-based green building design method further includes a timetable shape constraint step when simultaneously adjusting the timetable parameters in the two simulation engines at preset step sizes. Obtain the typical occupancy rate baseline curve, typical lighting utilization rate baseline curve, and typical equipment utilization rate baseline curve corresponding to the building type; After each timetable parameter combination is generated, the dynamic time warping algorithm is used to calculate the warped path distance between the adjusted timetable curve and the corresponding typical baseline curve. The regularized path distance is compared with a preset shape deviation tolerance threshold. Only when the regularized path distance is less than the shape deviation tolerance threshold is the set of timetable parameters used to re-execute the annual hourly energy consumption simulation.

[0011] Preferably, in the BIM-based green building design method, the shape deviation tolerance threshold is a graded threshold, and before comparing the dynamic time-normalized path distance with the shape deviation tolerance threshold, a threshold graded invocation step is also included: Based on the building type obtained from the IfcProject property, query the shape deviation tolerance level corresponding to the building type in the preset grading threshold mapping table; The hierarchical threshold mapping table stores the mapping relationship between office buildings and the first threshold, commercial buildings and the second threshold, and educational buildings and the third threshold, with the first threshold being less than the second threshold and less than the third threshold.

[0012] Preferably, the BIM-based green building design method further includes a spatial area-based fine-tuning step after the threshold-based grading call step. Extract the floor identifier and window orientation identifier of each IfcSpace. According to the preset regional fine-tuning rules, increase the shape deviation tolerance threshold by one level for IfcSpaces on the ground floor and facing the street, and decrease the shape deviation tolerance threshold by one level for IfcSpaces on the top floor. Increasing the threshold by one level means adjusting the current threshold to the next lower threshold with a larger value in the hierarchical threshold mapping table, and decreasing the threshold by one level means adjusting the current threshold to the next lower threshold with a smaller value. Replace the original shape deviation tolerance threshold of the corresponding IfcSpace with the adjusted hierarchical threshold for subsequent comparison with the regularized path distance.

[0013] The present invention has at least the following beneficial effects: 1. This invention automatically identifies the spatial function type of each IfcSpace through a semantic matching model. Combined with the building type, it automatically obtains personnel density, lighting power density, equipment power density, and corresponding schedules from a pre-set internal thermal disturbance parameter rule base, avoiding the tedious manual input of internal thermal disturbance parameters. Missing geometric information and thermal performance parameters of the building envelope are standardized and supplemented to ensure data consistency. A dual-engine parallel simulation and cross-validation are performed using a main energy consumption simulation engine and an auxiliary energy consumption simulation engine. Normalized mean error and root mean square error coefficient of variation are used as reliability criteria, ensuring the simulation results undergo reliability verification before output. When energy consumption exceeds the limit, the excess energy consumption can be automatically decomposed to each building envelope component and component-level structured modification instructions can be generated. These instructions are directly written into the BIM model through an application programming interface and trigger a new round of simulation, achieving automated closed-loop feedback from analysis to modification, reducing manual intervention and design iterations.

[0014] 2. In this invention, when the thermal performance parameters of a component have reached the upper limit of the performance of similar components in the high-performance component parameter library, but the energy consumption still exceeds the standard, the component parameters are automatically locked and the design strategy is switched to a passive approach. Ineffective adjustments to the thermal parameters of the building envelope itself are no longer made. By converting the excess energy consumption difference into an equivalent target value for solar radiation heat gain reduction, a quantitative mapping relationship is established between the excess energy consumption and the passive component parameters. This makes the determination of the window-to-wall ratio adjustment range or the size parameters of external shading components based on evidence, avoiding the blindness of trial and error based on experience. The corresponding adjustment scheme is automatically queried from the preset passive component parameter adjustment rule library, generating a structured modification instruction containing component identification and parameter values, and writing it into the BIM model. This allows the optimization process to automatically switch to the passive design optimization path after the building envelope performance reaches its limit, avoiding stagnation in the optimization process.

[0015] 3. In this invention, before the window-to-wall ratio modification command is written into the BIM model, the geometric boundary and visible light transmittance parameters of the space to which the window components belong are automatically extracted from the BIM model. Combined with typical meteorological year sky illuminance data, the percentage of total natural daylighting time is calculated, achieving automatic quantitative assessment of natural daylighting performance. When the daylighting index is lower than the green building daylighting compliance threshold, the window-to-wall ratio modification command is rejected while the original window-to-wall ratio parameters are retained, preventing damage to indoor natural daylighting quality due to a one-sided pursuit of reducing solar radiation heat gain. Simultaneously, after rejecting the window-to-wall ratio modification command, the system automatically searches the passive component parameter adjustment rule library and prioritizes the use of adjustable external shading component size parameters as an alternative. This leverages the flexible adjustment capability of adjustable shading components between reducing solar radiation heat gain and introducing natural light, achieving a balance between energy consumption optimization and daylighting assurance.

[0016] 4. In this invention, when retrieving and calling the adjustable external shading component size parameters, the location coordinates and geometric dimensions of window components in adjacent spaces on the same facade are automatically extracted from the BIM model. The overlapping area ratio is calculated through parallel projection, quantifying the spatial geometric relationship between adjacent window components. When the overlapping area ratio exceeds a preset shading threshold, the external shading component size parameters are reduced according to a reduction factor, ensuring that the final generated external shading component size reflects the effective shading requirements under the actual shading conditions of adjacent window components. This approach avoids the problem of unrepresented mutual shading effects caused by independent external shading designs for each window component, ensuring that the size parameters of the external shading component match the actual window layout spatial relationship of the building facade.

[0017] 5. In this invention, when multiple adjacent spaces on the same facade simultaneously generate new instructions for external shading components, a unified master-slave coordination processing order is established by extracting the window area parameters of each space and taking the space with the largest window area as the dominant space of that facade. The size parameters of the external shading components in the dominant space are determined first and used as the reference shading parameters for that facade. When generating instructions for external shading components in non-dominant spaces, the reference shading parameters are used as known boundary conditions for reduction calculations. This ensures that the size parameters of the external shading components in each space on the same facade are coordinated and determined under a unified reference framework, avoiding iterative conflicts such as mutual weakening or excessive superposition that may occur when each space adjusts independently.

[0018] 6. In this invention, when dual-engine cross-validation fails, the schedule parameters used to define personnel occupancy rate, lighting utilization rate, and equipment utilization rate in the two simulation engines are identified as the main source of uncertainty in the deviation. The schedule parameters of both engines are simultaneously adjusted within a preset step size and within the parameter percentage range. This process replaces the manual comparison and troubleshooting of parameter differences by iterating through multiple parameter combinations programmatically. The last set of schedule parameter combinations that meets the tolerance requirements, along with the corresponding energy consumption index data, is used as the final reliable output. This ensures that the self-calibration process has a clear operational path and reproducibility, reducing reliance on the professional experience of analysts.

[0019] 7. In this invention, a shape constraint mechanism is introduced during the self-calibration process, using the typical occupancy rate, lighting utilization rate, and equipment utilization rate benchmark curves corresponding to the building type as a reference. A dynamic time warping algorithm is used to calculate the warped path distance between the adjusted timetable curve and the typical benchmark curve. The degree of deviation of the overall curve shape is used as the criterion for judging the rationality of the parameter combination. Only when the warped path distance is less than the preset shape deviation tolerance threshold is the set of timetable parameter combinations used to re-execute the simulation calculation. This establishes a balance between mathematical convergence and the rationality of building function, avoiding parameter combinations that, while meeting the dual-engine tolerance requirements, result in severely distorted timetable curve shapes during the self-calibration process.

[0020] 8. In this invention, the shape deviation tolerance threshold is set as a graded threshold. For office buildings, commercial buildings, and educational buildings, a first threshold, a second threshold, and a third threshold are set in ascending order, respectively. This ensures that the strictness of the shape constraint matches the actual requirements of different building types for the regularity of their usage patterns. Before comparing the regularized path distance with the threshold, the corresponding level of threshold is automatically called based on the building type obtained from the IfcProject attribute. This avoids the problem of insufficient adaptability caused by using a uniform fixed threshold. It ensures that the calibration results for building types with strong regularity, such as office buildings, accurately reflect the actual usage characteristics, while providing reasonable parameter adjustment space for building types with greater flexibility, such as commercial and educational buildings.

[0021] 9. In this invention, based on the tiered threshold call, the shape deviation tolerance threshold is further fine-tuned according to the floor identification and window orientation of each IfcSpace. The threshold level is raised by one level for the ground floor street-facing facade to accommodate the high flexibility of personnel activity in this area due to its external service functions; the threshold level is lowered by one level for the top floor space to match the more regular usage patterns of the top-floor office area. Through this refined fine-tuning by spatial region, spaces in different locations within the same building can adopt a shape constraint strictness that matches their actual usage patterns, improving the adaptability of the calibration results to the actual conditions of different areas while maintaining self-calibration convergence.

[0022] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Detailed Implementation

[0023] The present invention will now be described in further detail so that those skilled in the art can implement it based on the description.

[0024] It should be understood that terms such as “having,” “comprising,” and “including” as used herein do not exclude the presence or addition of one or more other elements or combinations thereof.

[0025] According to one embodiment of the present invention, a BIM-based green building design method is provided. First, a BIM model of the target building is constructed. This BIM model is stored as an IFC format file and can be created using mainstream BIM modeling software that supports IFC standard output. The model needs to include the geometric information of building components, specifically including the three-dimensional dimensions and spatial location of the building envelope, such as exterior walls, windows, roof, and floors. Thermal performance parameters of the building envelope can include the heat transfer coefficient of exterior walls, the heat transfer coefficient of exterior windows, the solar heat gain coefficient of exterior windows, and the heat transfer coefficient of the roof. Building material attribute information can include the material name, thickness, density, specific heat capacity, and thermal conductivity of each structural layer of the walls. Spatial objects are expressed as IfcSpace instances. Each IfcSpace instance corresponds to an independent functional space within the building, and its attributes should include the name and area information of that space. Taking a twelve-story office building as an example, the IFC file can contain approximately two hundred IfcSpace instances, corresponding to space types such as offices, meeting rooms, corridors, and restrooms. The name attribute of each IfcSpace can be filled in according to design conventions, such as "Standard Office," "Open Office Area," or "Small Meeting Room."

[0026] Geometric information and thermal performance parameters of the building envelope are extracted from IFC format files. Geometric information extraction is achieved by parsing IfcProduct and its subclass entities in the IFC file, converting the extracted geometric information into gbXML geometric description data. During the conversion process, components such as IfcWall, IfcWindow, IfcSlab, and IfcRoof are mapped to Surface elements in gbXML, and the vertex coordinates, area, orientation, and other geometric attributes of each Surface are calculated from the IFC geometric expression. Thermal performance parameters of the building envelope are extracted by reading IfcMaterial and its associated attribute set, and then mapping the thermal performance parameters to material attribute fields in gbXML format, specifically to fields such as Conductivity, U-value, and SHGC of the Material element. When geometric information of a component is missing in the IFC file, such as a wall lacking thickness data, the spatial location data and component type identifier of the component can be extracted from the IFC format file. Based on the spatial location data and component type identifier, corresponding supplementary geometric parameters are searched from a pre-defined building component information database and written into the gbXML geometric description data. The pre-set building component information database can be compiled according to common building types and component types, storing default geometric parameters such as standard exterior wall thickness, standard window sill height, and standard floor height. When a missing item is detected in the thermal performance parameters of the building envelope, such as a window lacking solar heat gain coefficient, the database of pre-set building material thermal parameters is retrieved based on the building material attribute information. This database can be compiled according to national building standard design atlases and general building material thermal parameters to obtain the corresponding default values ​​of the building envelope thermal performance parameters, and these default values ​​are then filled into the material attribute fields in gbXML format.

[0027] The name and description attributes of IfcSpace instances are extracted, and a pre-trained semantic matching model is used to determine the spatial function type of each IfcSpace. The semantic matching model can employ a Word2Vec word vector model trained on an architectural space corpus, which collects spatial naming texts from various architectural design specifications as training data. The model maps the name attributes of IfcSpace to a high-dimensional semantic space and calculates its cosine similarity with preset standard spatial function type labels. These preset standard spatial function type labels can include offices, meeting rooms, corridors, lobbies, restrooms, equipment rooms, storage rooms, etc. The label with the highest similarity score above 0.7 is selected as the function inference result for that space. When the highest similarity score is below 0.7, it can be labeled as "other functional spaces" and a preset general internal thermal perturbation parameter is used. Simultaneously, the building type is obtained from the IfcProject attribute. IfcProject is the root entity of the IFC model hierarchy, and its building type attribute field can be read as values ​​such as "office building" or "commercial building". When this attribute is empty, the area ratio of each spatial function type in the inference results of each IfcSpace function is calculated, and the spatial function type with the largest area ratio is determined as the dominant function type of the building. Then, the building type is determined according to the preset mapping relationship table between the dominant function type of the building and the overall building type. For example, when the area ratio of office space exceeds 50%, the building type is determined to be an office building.

[0028] Indexed by building type and space function type, the system retrieves the personnel density, lighting power density, equipment power density, and corresponding schedules for each IfcSpace from a pre-defined internal thermal disturbance parameter rule base. The internal thermal disturbance parameter rule base can be compiled based on the correspondence between space type and load density in ASHRAE (American Society of Heating, Refrigerating and Air-Conditioning Engineers) Standard 90.1, Public Building Energy Efficiency Design Standard GB 50189, and Building Lighting Design Standard GB 50034. Taking the "office" space type in an office building as an example, the personnel density can be taken as 0.1 people per square meter, the lighting power density as 9 watts per square meter, and the equipment power density as 15 watts per square meter. The office attendance rate schedule for weekdays can be set as follows: 0:00-7:00, 7:00-8:00, 20%, 8:00-12:00, 90%, 12:00-13:00, 30%, 13:00-17:00, 90%, 17:00-19:00, 30%, 19:00-24:00; 0% for rest days. The lighting and equipment usage schedules can be set accordingly based on the attendance rate curve. The obtained parameters should be written into the corresponding `PeopleHeatGain`, `LightPowerPerArea`, `EquipPowerPerArea`, and `Schedule` attributes of the `Space` element in the gbXML format data, according to the element tags defined in the gbXML Schema.

[0029] The completed gbXML data, including geometric information, thermal parameters, and internal thermal disturbance parameters, is imported into the main energy consumption simulation engine and an auxiliary energy consumption simulation engine based on different mathematical mechanisms. The main energy consumption simulation engine can use EnergyPlus (a building energy simulation software developed by the U.S. Department of Energy), whose core is based on the heat balance method to solve the heat and moisture transfer process in building rooms. The auxiliary energy consumption simulation engine can use an energy consumption simulation model built on the Modelica (an object-oriented physical modeling language) Buildings Library, whose core is based on the thermal network method, employing a different mathematical and physical modeling path than EnergyPlus. Both simulation engines use the exact same gbXML input file and the exact same hourly meteorological data file for a typical meteorological year, independently performing hourly energy consumption simulations throughout the year. The meteorological data file can be selected from the Chinese standard year meteorological database, containing hourly dry-bulb temperature, wet-bulb temperature, direct solar radiation, diffuse radiation, wind speed, and wind direction for the project location. The main energy consumption simulation engine outputs the first energy consumption index data, and the auxiliary energy consumption simulation engine outputs the second energy consumption index data. The energy consumption index data can be the annual heating and cooling energy consumption per unit building area.

[0030] Calculate the Normalized Mean Bias Error (NMBE) and the Coefficient of Variation of Root Mean Square Error (CV(RMSE)) between the first and second energy consumption index data. The NMBE is calculated as the ratio of the monthly energy consumption difference between the two simulation results to the monthly average energy consumption. The CV(RMSE) is calculated as the ratio of the root mean square of the monthly energy consumption difference to the average energy consumption of the two simulation results. When the absolute value of the NMBE is less than 15% and the CV(RMSE) is less than 20%, the deviation between the two simulation engines is considered to be within a preset engineering tolerance threshold, and the first energy consumption index data is used as the final energy consumption index data. If the above tolerance conditions are not met, model self-calibration is performed. The specific self-calibration process is as follows: Extract the time schedule parameters used to define the occupancy rate, lighting utilization rate, and equipment utilization rate from the main energy consumption simulation engine and the auxiliary energy consumption simulation engine respectively. Adjust the time schedule parameters in the two simulation engines synchronously within the percentage range of their respective time schedule parameters, with a step size of 5%. For example, adjust the peak occupancy rate of 90% to between 85% and 95% in steps of 5%. Adjust the lighting utilization rate and equipment utilization rate accordingly. Generate multiple sets of time schedule parameter combinations and re-execute the annual hourly energy consumption simulation until the calculated normalized mean error and root mean square error coefficient of variation both meet the above tolerance requirements. Use the last set of time schedule parameter combinations that meet the tolerance requirements and its corresponding first energy consumption index data as the final reliable simulation input and output results.

[0031] The final energy consumption index data is compared with the preset green building energy consumption threshold range. For Class A office buildings in cold regions, the upper limit of the green building energy consumption threshold range can be set at 45 kWh per square meter per year for heating and cooling. This threshold can be determined by referring to the requirements for the thermal performance of the building envelope in the current green building evaluation standards. When the final energy consumption index data exceeds this threshold range, the excess energy consumption is decomposed into each building envelope component to form a list of key components. The decomposition method can be based on the area, heat transfer coefficient, and temperature difference-driven load contribution rate of each component to calculate the contribution ratio of each component to the total excess energy consumption, and components with a contribution ratio exceeding 10% are selected to generate a list of key components. Based on a pre-defined high-performance component parameter library, structured modification instructions at the component level are generated. This library stores alternative parameter values ​​for different performance levels according to component type. For example, the heat transfer coefficient of external walls can be stored in tiers of 0.20, 0.25, 0.30, 0.35, and 0.40 Kelvin per square meter, while the heat transfer coefficient of external windows can be stored in tiers of 1.0, 1.5, and 2.0 Kelvin per square meter. For each component in the key component list, its current thermal performance parameters are compared with the parameter range of similar components in the high-performance component parameter library. A parameter value one level higher than the current value is selected as the target parameter value, forming a structured modification instruction containing a unique component identifier, the name of the parameter to be modified, the current parameter value, and the target parameter value. Then, the modification instructions are traversed through the BIM modeling software's application programming interface. The target component in the BIM model is located based on the unique component identifier, and the target parameter value is written into the component's attribute field. After all modification instructions have been executed, a new round of gbXML file export and building energy consumption simulation calculations is automatically triggered.

[0032] Compared to common BIM-based energy consumption simulation methods, which typically require manual identification of spatial functions within the BIM model and manual input of internal thermal disturbance parameters, relying on operator experience to supplement missing geometric and thermal information, and using simulation results directly from a single simulation engine for design decisions, this approach suffers from numerous issues including excessive manual intervention, inconsistent parameter completion standards, lack of cross-validation of simulation results, and incomplete design feedback. This solution automatically infers spatial function types through semantic matching models and automatically completes internal thermal disturbance parameters using a rule base. A dual-engine cross-validation mechanism verifies the reliability of simulation results. When energy consumption exceeds limits, it automatically generates component-level structured modification instructions and writes them into the BIM model, triggering a new round of simulation. This achieves end-to-end automated data flow from BIM model to energy consumption analysis and design modification, reducing manual intervention and improving the consistency of parameter completion and the reliability of simulation results.

[0033] According to another embodiment of the present invention, in the provided BIM-based green building design method, during the process of automatically generating component-level structured modification instructions and writing them into the BIM model, when the program detects that the current thermal performance parameters of a certain component in the key component list have reached the upper limit of the performance of the same type of component in the preset high-performance component parameter library, and the final energy consumption index data still exceeds the green building energy consumption threshold range, it indicates that simply relying on improving the thermal performance of the building envelope itself can no longer further meet the energy-saving requirements. At this time, it is necessary to shift the optimization direction from the thermal performance of the building envelope itself to the passive design strategy of the entire building, by adjusting the window-to-wall ratio of the building facade or adding external shading components to reduce solar radiation heat gain, thereby reducing the building's air conditioning cooling load from the source.

[0034] When a component is detected to have reached its performance limit, its parameters are first locked to the performance limit value to prevent repeated and invalid adjustments in subsequent iterations. Next, the building orientation and facade identifier corresponding to the component are extracted. The building orientation can be obtained from the IfcBuilding entity orientation attribute in the BIM model, or it can be calculated based on the true north direction information in IfcSite combined with the component's external normal direction. The facade identifier can be categorized based on the angle between the component's external normal and the four basic directions: east, south, west, and north. For example, components whose external normal direction falls within the range of 45 degrees southeast to 45 degrees southwest are classified as south-facing facades.

[0035] After determining the orientation and facade of a component, the excess energy consumption difference corresponding to the green building energy consumption threshold range in step six is ​​converted into an equivalent solar radiation heat gain reduction target value. The specific conversion method is as follows: the equivalent solar radiation heat gain reduction equals the excess energy consumption difference divided by the comprehensive energy efficiency ratio of the air conditioning system, where the comprehensive energy efficiency ratio of the air conditioning system can be taken as 3.0 according to the annual comprehensive performance coefficient of the air conditioning system specified in the current public building energy-saving design standards. For example, when a south-facing office's annual energy consumption per unit area exceeds the threshold by 5 kWh per square meter due to the heat transfer coefficient of its windows reaching the performance limit, the calculated equivalent solar radiation heat gain reduction target value is 5 divided by 3.0, which is approximately 1.67 kWh per square meter per year. This value represents the amount of solar radiation heat gain that needs to be reduced from the facade by adjusting the window-to-wall ratio or adding external shading components.

[0036] Based on the target value for solar radiation heat gain reduction, the corresponding window-to-wall ratio adjustment range or external shading component size parameters are retrieved from the preset passive component parameter adjustment rule library. The passive component parameter adjustment rule library can be established separately according to different building climate zones, different orientations, and different window types. Taking a south-facing facade in a cold region as an example, the rule library can store the following correspondences: When the target value for solar radiation heat gain reduction is within the range of 0 to 3 kWh / m² / year, it is recommended to reduce the window-to-wall ratio by 0.02 to 0.05 from the current level, or to add a horizontal fixed external shading panel with a projecting length of 0.3 to 0.6 meters; when the target value is within the range of 3 to 6 kWh / m² / year, it is recommended to reduce the window-to-wall ratio by 0.05 to 0.10, or to add a horizontal fixed external shading panel with a projecting length of 0.6 to 1.0 meters. For east-west facing facades, due to the lower solar altitude angle, the corresponding recommended measures in the rule library can focus on adding vertical shading components or adjustable external shading louvers.

[0037] Based on the query results, generate window-to-wall ratio modification instructions or external shading component addition instructions, each containing a unique identifier for the window component and an exterior facade identifier. The window-to-wall ratio modification instruction includes the unique identifier of the window component being modified in the BIM model, the current window-to-wall ratio, the target window-to-wall ratio, and the corresponding exterior facade identifier. The external shading component addition instruction includes the associated window component's unique identifier, the shading component type (e.g., horizontal shading panel or adjustable louver), the shading component's geometric parameters (e.g., cantilever length and width), and the offset of the shading component's installation position relative to the top or side of the window. The generated new instructions are written into the BIM model via the BIM modeling software's application programming interface (API). Specifically, logical associations between the newly added external shading components and the corresponding window components can be established using entities such as IfcRelAssignsToProduct. This automatically triggers a new round of gbXML file export and building energy consumption simulation calculations, entering the next round of analysis and verification cycles.

[0038] In existing BIM-based energy consumption simulation and optimization processes, when the thermal performance of building envelope components is optimized to the upper limit of the product library but energy consumption still fails to meet standards, the common approach is to manually adjust the window-to-wall ratio or add shading components after manual judgment, and then re-export the model for simulation. This entire process lacks an automated quantitative conversion mechanism and a direct correlation with the original energy consumption exceedance. The targeted nature and efficiency of the adjustments rely heavily on the designer's personal experience. The proposed solution establishes a quantitative conversion relationship between the excess energy consumption difference and the target value of equivalent solar radiation heat gain reduction. Using a pre-defined passive component parameter adjustment rule library as a bridge, it directly maps the energy consumption exceedance to the window-to-wall ratio adjustment range or the external shading component size parameters. This achieves automatic generation and writing of passive component modification instructions from energy consumption exceedance signals, reducing manual intervention.

[0039] According to another embodiment of the present invention, in the provided BIM-based green building design method, before generating the window-to-wall ratio modification instruction and writing it into the BIM model, a natural lighting compliance verification step needs to be performed to prevent the reduction of the window-to-wall ratio from having an adverse effect on the indoor natural lighting performance, and to avoid a situation where energy consumption optimization and lighting performance are compromised.

[0040] The specific process for verifying the compliance of natural lighting is as follows: Extract the geometric boundary information of the space to which the window component belongs from the BIM model. This information can include the length, width, clear height of the space, and the height of the reference working plane. The visible light transmittance parameter can be extracted from the glass material properties associated with the window component in the IFC file. If the IFC file does not store the visible light transmittance parameter, a default value can be obtained from a pre-defined library of commonly used glass optical parameters based on the type of window component. For example, the visible light transmittance of ordinary double-glazed windows can be 0.78, and that of Low-E double-glazed windows can be 0.62. Sky illuminance data from typical meteorological year files can be obtained from the EnergyPlus meteorological files of the project location. These files contain hourly direct and diffuse solar illuminance data for the entire year. Using the Perez all-sky illuminance model, the hourly natural light illuminance values ​​at each calculation point on the reference working plane can be calculated.

[0041] When calculating the percentage of total natural daylight time on the reference working plane, the height of the reference working plane can be taken as 0.75 meters, and the calculation grid spacing can be taken as 0.5 meters. The percentage of total natural daylight time is calculated as the percentage of hours with an illuminance of at least 300 lux on the reference working plane during the entire year's usage time, which can be taken as 8:00 AM to 6:00 PM, totaling 3650 hours per year. During the calculation, a rapid estimation can be performed using the daylight factor method combined with sky illuminance data, or a precise point-by-point calculation can be performed using natural daylight simulation programs such as Radiance (a ray-tracing-based natural daylight simulation software).

[0042] The preset threshold for green building daylighting compliance can be set according to the scoring rules for indoor natural daylighting of office buildings in the current green building evaluation standard GB / T 50378, and the threshold can be set to 60%. When the calculated percentage of total natural daylighting time is less than 60%, it indicates that the natural daylighting performance of the space under the current window-to-wall ratio is still on the verge of compliance. Further reducing the window-to-wall ratio will cause the natural daylighting to fail to meet the green building evaluation requirements. At this time, the system rejects the window-to-wall ratio modification command, retains the original window-to-wall ratio parameter, and searches the passive component parameter adjustment rule library to preferentially call the adjustable external shading component size parameter corresponding to the window component.

[0043] The reason for prioritizing adjustable external shading components over fixed ones is that adjustable components, such as movable louvers and retractable blinds, can be deployed during periods of strong solar radiation to block heat gain and retracted during cloudy days or when natural light is needed to allow natural light in. This has a smaller impact on the percentage of total natural lighting time, allowing for both reduction of solar heat gain and improved natural lighting performance. The dimensional parameters of adjustable external shading components can include the width of the louver blades, the blade spacing, the adjustable range of the blade tilt angle, and the overall cantilever length of the shading component. These parameters can be automatically matched from the passive component parameter adjustment rule library based on the window component size and orientation. After matching, a new external shading component instruction is generated and written into the BIM model. This instruction includes a unique identifier for the associated window component, a shading component type identifier of adjustable external shading louvers, the geometric dimensions of the shading component, and the installation position offset.

[0044] In common BIM-based energy consumption simulation and optimization processes, window-to-wall ratio adjustments are typically made manually by designers based on energy consumption simulation results. These modifications do not automatically trigger natural lighting verification. Lighting performance verification is usually conducted separately by professionals using independent software after energy consumption optimization is complete. If insufficient lighting is found, the window-to-wall ratio needs to be readjusted and energy consumption simulation repeated, resulting in multiple iterations. The proposed solution makes natural lighting compliance verification a prerequisite for window-to-wall ratio modification commands to take effect. Verification and rejection are automatically completed before the command is written into the BIM model. After rejection, adjustable external shading components are prioritized as alternatives. This allows energy consumption optimization and lighting assurance to be completed collaboratively in the same process, reducing design iterations.

[0045] According to another embodiment of the present invention, in the provided BIM-based green building design method, during the process of retrieving and calling the adjustable external shading component size parameters from the passive component parameter adjustment rule library, it is necessary to consider the mutual shading relationship between window components in adjacent spaces on the same facade. When a new external shading component instruction is generated for the window components of the currently verified space, the newly added external shading component may, while blocking solar radiation, also cause cross-space shading of window components in adjacent spaces, affecting the natural lighting or solar radiation heat gain calculation of adjacent spaces. Similarly, external shading components that have been installed or are planned to be installed in adjacent spaces may also cause shading of the window components in the currently verified space. To solve this cross-space shading problem, a cross-space shading coordination step needs to be performed when retrieving and calling the external shading component size parameters.

[0046] The cross-space occlusion coordination process first extracts the position coordinates and geometric dimensions of window components in adjacent spaces on the same facade as the current window component from the BIM model. Determining whether a window component is on the same facade can be based on the outward normal direction of each window component; window components with an angle less than 15 degrees between their outward normal directions are grouped into the same facade. The position coordinates of a window component can be the three-dimensional coordinates of its geometric center point, and the geometric dimensions can include the width and height of the window. Adjacent spaces can be identified based on the IfcRelSpaceBoundary relationship to determine the adjacency between each IfcSpace, or by calculating the spatial distance between the geometric center points of each window component. Two window components with a distance less than a preset adjacency threshold, such as 3 meters, can be considered as window components in adjacent spaces.

[0047] After extracting the window component information of adjacent spaces, the window components of the adjacent spaces are projected parallel to the current verification space. The direction of the parallel projection is the opposite of the outward normal direction of the facade, that is, the direction perpendicular to the facade towards the building interior. During projection, the geometric outline of the adjacent window components is extended along the projection direction to the facade plane where the window components of the current verification space are located, and the overlap area between the projected area and the geometric outline of the window components of the current verification space is calculated. The overlap area ratio is calculated by dividing the projected overlap area by the area of ​​the window components of the current verification space. This ratio reflects the proportion of the light opening of the current window that may be blocked by the external shading components of the adjacent space.

[0048] The preset occlusion threshold can be set to 25%. When the overlap area exceeds 25%, it indicates that the adjacent window components significantly obstruct the current window component being checked, requiring a reduction in the size of the external shading component in the current check space. The reduction calculation can be performed by subtracting the retrieved adjustable external shading component size parameters according to a preset reduction coefficient. The reduction coefficient can be set in segments based on the overlap area percentage. For example, when the overlap area percentage is between 25% and 50%, the reduction coefficient can be 0.8; when the overlap area percentage is between 50% and 75%, the reduction coefficient can be 0.6; and when the overlap area percentage exceeds 75%, the reduction coefficient can be 0.4. Taking the cantilever length of the adjustable external shading louver as an example, if the initial cantilever length retrieved from the rule base is 0.8 meters, after reduction by a coefficient of 0.8, the reduced cantilever length is 0.64 meters. The reduced dimensional parameters are used to generate a new instruction for the external shading component. This instruction replaces the original retrieved parameter instruction and is written into the BIM model.

[0049] When the overlap area does not exceed 25%, the retrieved adjustable external shading component size parameters are not reduced, and the new external shading component instruction is generated directly using the size parameters retrieved from the rule base. The cross-space shading coordination step only affects the final value of the external shading component size parameters in the current verification space, and does not change the already determined window component geometric parameters and external shading component parameters in adjacent spaces. The parameters of adjacent spaces are independently coordinated and calculated when they are each verification spaces.

[0050] In common BIM-based energy consumption simulation and external shading design processes, the selection and size determination of external shading components are usually done independently. The shading components for each room or window are specified individually by the designer, without considering the impact of the geometric relationship between adjacent window components on the shading effect. When the external windows of adjacent rooms are close together, the independently installed external shading components may obstruct or overlap each other, leading to a difference between the actual shading effect and the design expectation. The above solution, when performing external shading component size parameter retrieval, automatically extracts the geometric information of adjacent window components on the same facade and performs parallel projection analysis. It then reduces the external shading component size based on the proportion of overlapping shading area, achieving automatic quantification and coordination of the cross-space shading impact. This ensures that the final generated external shading component size reflects the actual spatial relationship between adjacent window components.

[0051] According to another embodiment of the present invention, in the provided BIM-based green building design method, when multiple adjacent spaces on the same facade simultaneously generate new instructions for external shading components, the independent execution of cross-space shading coordination steps by each space may lead to inconsistent parameter adjustment directions, or even situations where the external shading components of two adjacent spaces weaken or overlap each other, resulting in iterative oscillations. To solve this problem, a coordination control mechanism needs to be established on the facade to unify the generation order and benchmark of external shading component parameters for each adjacent space on the facade.

[0052] The main control step for adjacent space coordination first extracts the window area parameters of all spaces on the facade to be added as external shading components from the BIM model. The extraction scope is limited to all spaces on the same facade that simultaneously generated instructions to add external shading components in this simulation iteration. The criteria for determining the same facade are consistent with the cross-space shading coordination step, that is, window components with an angle of less than 15 degrees in the outward normal direction are classified as the same facade. The window area parameters can be directly extracted from the geometric expression associated with IfcWindow, or read from the area field of the corresponding Surface element in the gbXML geometric description data.

[0053] After extraction, all spaces to be added with external shading components are sorted by window area from largest to smallest, and the space with the largest window area is determined as the dominant space of the facade. The basis for selecting the space with the largest window area as the dominant space is that the larger the window area, the greater the solar radiation heat entering the room through that window, and the greater the weight of the space's impact on the overall floor's energy consumption. Using its external shading component parameters as a benchmark can make the shading coordination of the entire facade more conducive to reducing overall energy consumption.

[0054] Once the dominant space is determined, the command to add external shading components for that space is executed first. This command retrieves the original parameters from the passive component parameter adjustment rule base or processes them separately through a cross-space shading coordination step, without any additional coordination or reduction. After execution, the dimensions of the external shading components in the dominant space are used as the baseline shading parameters for that facade. These baseline shading parameters may include the projecting length of the external shading components, the width and spacing of the louvers, and the louver tilt angle range.

[0055] When generating a new instruction for external shading components in a non-dominant space, the baseline shading parameters are substituted as known boundary conditions into the cross-space shading coordination step for reduction calculation. Specifically, when performing the cross-space shading coordination step on window components in a non-dominant space, the window components and external shading components already determined in the dominant space are considered as existing shading sources in adjacent spaces. A reduction coefficient is calculated based on the overlap area ratio, and the size parameters of the external shading components retrieved from the rule base in the non-dominant space are reduced. The reduced size parameters serve as the final parameters for the new instruction for external shading components in a non-dominant space. When multiple non-dominant spaces exist, they are processed sequentially in descending order of window area. When performing the reduction calculation for each non-dominant space, the external shading component parameters of both the dominant space and the already processed non-dominant spaces are substituted as known boundary conditions.

[0056] Taking the south-facing facade of a standard floor of an office building as an example, there are three adjacent office spaces with window areas of 6 square meters, 4 square meters, and 3 square meters, respectively. In this simulation iteration, all three spaces simultaneously generated instructions to add external shading components. The main control step identified the space with a window area of ​​6 square meters as the dominant space and prioritized executing its instruction to add external shading components. Assuming that the adjustable external shading louver extension length retrieved from the rule base is 0.8 meters, then 0.8 meters becomes the baseline shading extension length for this facade. Subsequently, the non-dominant space with a window area of ​​4 square meters was processed. The overlapping area ratio of its window components was calculated by projecting them towards the window components of the dominant space. If the overlapping area ratio is 30%, the reduction factor is taken as 0.8, and the extension length of the external shading louver in this space is reduced to 0.64 meters. Finally, the non-dominant space with a window area of ​​3 square meters was processed. Considering the shading effects of both the dominant space and the processed non-dominant space, the reduction factor and the final extension length were determined after comprehensively calculating the overlapping area ratio.

[0057] In existing BIM-based energy consumption simulation and shading design processes, external shading components for multiple adjacent rooms on the same facade are typically specified or generated independently by designers, lacking a coordination mechanism between rooms. When multiple rooms adjust their external shading components independently due to excessive energy consumption, interference in shading effects or inconsistencies in component dimensions may occur, requiring designers to repeatedly coordinate based on experience. The proposed solution establishes a dominant space determination mechanism and a master-slave coordination sequence on the same facade. Using the space with the largest window area as the benchmark, the external shading component parameters for non-dominant spaces are reduced under the constraint of the benchmark shading parameters. This ensures that the external shading component parameters for all spaces on the same facade form a coordinated overall solution, reducing iterative conflicts that may arise from independent adjustments.

[0058] According to another embodiment of the present invention, in the provided BIM-based green building design method, when the normalized average deviation or root mean square error coefficient of variation between the first energy consumption index data and the second energy consumption index data output by the main energy consumption simulation engine and the auxiliary energy consumption simulation engine is not within the respective preset engineering tolerance threshold range, it indicates that there is a large deviation in the calculation results of the two simulation engines, and the input parameters of one or both simulation engines may be biased. At this time, the first energy consumption index data cannot be directly used as the final energy consumption index data. It is necessary to perform a model self-calibration step to correct the simulation input parameters so that the output results of the two engines tend to converge.

[0059] The starting point for model self-calibration is that, given the geometric information and thermal parameters of the building envelope have been determined through the aforementioned completion steps, the primary source of uncertainty in the calculation results of the two simulation engines is usually the timetable parameters of internal thermal disturbances. Although the primary and auxiliary energy consumption simulation engines differ in their mathematical and physical models, their calculation results should fall within the preset engineering tolerance range when both use the exact same timetable parameters. Therefore, the core idea of ​​self-calibration is to find a parameter combination that allows the deviation between the calculation results of the two simulation engines to meet the tolerance requirements by synchronously adjusting the timetable parameters of the two simulation engines.

[0060] The self-calibration process first extracts the schedule parameters used to define occupancy rate, lighting utilization rate, and equipment utilization rate from both the primary and secondary energy consumption simulation engines. The schedule parameters for the primary energy consumption simulation engine can be extracted from data fields of objects such as Schedule:Compact or Schedule:Year from its input data file, such as an EnergyPlus .idf file. The schedule parameters for the secondary energy consumption simulation engine can be extracted from the corresponding Modelica model's input data file as a schedule array. The extracted schedule parameters are expressed as hourly percentage values ​​in their respective simulation engines, ranging from zero to one hundred percent.

[0061] After extraction, the timetable parameters in both simulation engines are synchronously adjusted within the percentage range of their respective timetable parameters, using a preset step size. The preset step size can be 5%, and the percentage range can be set as the fluctuation range centered on the current timetable parameter value. The upper and lower limits of the fluctuation range can be ±20% of the current value and must not exceed the physical boundary between 0 and 100%. Synchronous adjustment means that for the same timetable type, such as the occupancy rate timetable, both the primary energy consumption simulation engine and the auxiliary energy consumption simulation engine are modified in the same direction and with the same adjustment magnitude, ensuring that both engines use the same timetable parameters under any set of parameter combinations.

[0062] Taking office buildings as an example, if the peak value of the current model's weekday attendance rate is 90%, the peak value of lighting utilization is 80%, and the peak value of equipment utilization is 70%, then during the self-calibration process, the fluctuation range of each parameter is set as follows: the peak attendance rate is adjusted in 5% increments between 70% and 100%, the peak lighting utilization rate is adjusted in 5% increments between 60% and 100%, and the peak equipment utilization rate is adjusted in 5% increments between 50% and 90%. The adjustment combination of the three schedule parameters constitutes a three-dimensional parameter space. The self-calibration program iterates through the parameter combinations one by one in the order of attendance rate, lighting utilization rate, and equipment utilization rate. After generating a set of schedule parameter combinations each time, the set of parameters is written into the input data files of the two simulation engines, and the hourly energy consumption simulation for the whole year is re-executed, outputting a new round of first energy consumption index data and second energy consumption index data respectively.

[0063] Calculate the normalized mean error (NME) and root mean square error (RMSE) coefficient of variation for the next round, and determine whether both fall within their respective preset engineering tolerance thresholds. The engineering tolerance threshold for the NME can be set to an absolute value less than 15%, and the engineering tolerance threshold for the RMSE coefficient of variation can be set to less than 20%. When the tolerance conditions are met, stop the iteration and use the set of timetable parameter combinations and their corresponding first energy consumption index data as the final reliable simulation input and output results. If the tolerance conditions are still not met after iterating through all parameter combinations, select the parameter combination that minimizes the combined values ​​of the NME and RMSE coefficient of variation as the optimal approximation result, and record this in the output report for the designer's reference.

[0064] In common dual-engine energy consumption simulation verification processes, when the calculation results of the two simulation engines deviate beyond the tolerance range, the usual practice is for analysts to manually check the differences in the input parameters of the two models, comparing geometric models, thermal parameters, time parameters, meteorological data, etc., to identify the source of the discrepancy, manually correct it, and rerun the simulation. This entire process is time-consuming and relies heavily on the analyst's professional experience. The proposed solution identifies the time parameters as the main source of uncertainty in the deviation. By programmatically traversing and synchronously adjusting the time parameters of the two simulation engines, it automatically finds the parameter combinations that can bring the results of the two engines to converge, and directly uses the parameter combinations that meet the tolerance requirements as the final reliable simulation input and output, reducing the workload of manual checking and repeated adjustments.

[0065] According to another embodiment of the present invention, in the provided BIM-based green building design method, when the timetable parameters in the main energy consumption simulation engine and the auxiliary energy consumption simulation engine are adjusted synchronously at preset step sizes, the self-calibration process, which aims purely at mathematical convergence, may generate a set of timetable parameter combinations that, while meeting the dual-engine tolerance requirements, are unreasonable in terms of building function. For example, after self-calibrating the timetable of an office building, situations may arise where the peak occupancy rate occurs at 2:00 AM and drops to zero at noon, significantly deviating from actual usage patterns. Although such parameter combinations numerically make the calculation results of the two simulation engines more consistent, the energy consumption index data derived from these parameter combinations cannot reflect the actual operating characteristics of the building because they distort the building usage scenario they describe. To prevent such distorted parameter combinations from occurring during the self-calibration process, a timetable shape constraint step needs to be introduced after each timetable parameter combination is generated.

[0066] The timetable shape constraint step first obtains the typical occupancy rate baseline curve, typical lighting utilization rate baseline curve, and typical equipment utilization rate baseline curve corresponding to the building type. The building type can be obtained from the IfcProject attribute or inferred from the aforementioned semantic matching model based on the area ratio of each space function type. The typical baseline curve can be derived from the recommended timetable curves classified by building type in the appendix of ASHRAE Standard 90.1, or from the occupancy, equipment, and lighting operation timetables of various public buildings given in the appendix of the Public Building Energy Conservation Design Standard GB50189, or from publicly available typical building operation mode datasets in the field of building energy consumption simulation. Taking the typical occupancy rate baseline curve of an office building as an example, its weekday curve shape characteristics are as follows: the occupancy rate is close to zero from 0:00 to 6:00, slowly rises from 6:00, rapidly rises to a peak of about 80% from 7:00 to 8:00, remains near the peak from 8:00 to 12:00, slightly decreases to about 30% from 12:00 to 13:00 due to lunch break, recovers to the peak from 13:00 to 17:00, rapidly decreases to zero after 17:00, and remains basically zero at night. The shape characteristics of the typical lighting utilization rate baseline curve and the typical equipment utilization rate baseline curve are similar to those of the personnel occupancy rate curve, but the peak value and the peak-to-valley difference are different. The peak value of lighting utilization rate can be taken as 80%, and the peak value of equipment utilization rate can be taken as 70%.

[0067] After generating a set of timetable parameter combinations in each self-calibration process, the dynamic time warping algorithm is used to calculate the warped path distance between the adjusted timetable curve and the corresponding typical reference curve. The dynamic time warping algorithm is an algorithm for measuring the similarity between two time series. It finds the optimal matching path between two series through non-linear time axis warping, exhibiting good robustness to local stretching, compression, and phase shifts of the curve on the time axis. Taking the occupancy rate timetable as an example, the adjusted hourly occupancy rate curve for weekdays is considered the series to be compared, and the typical office building's weekday occupancy rate reference curve is considered the reference series. Both series have twenty-four discrete points, each corresponding to the hourly average from 00:00 to 23:00 on that day. The dynamic time warping algorithm calculates the Euclidean distance matrix between each point in the two series. On this matrix, a warped path with the minimum cumulative distance is searched using dynamic programming. The sum of the distances at each point on the path is the warped path distance. The absolute value of the warped path distance is greatly affected by the absolute value of the curve peak; therefore, a normalized form, the warped path distance divided by the product of the series length and the peak value, is used to make this metric universal for different building sizes. The calculation method for the regularized path distance of the lighting utilization rate curve and the equipment utilization rate curve is the same as that for the personnel occupancy rate curve.

[0068] The calculated regularized path distance is compared with a preset shape deviation tolerance threshold. The shape deviation tolerance threshold, when no tiered thresholds are used, can be set to 0.3, a value determined based on statistical experience of typical fluctuation ranges in the shape of office building timetables. When the regularized path distance between the adjusted timetable curve and the typical baseline curve is less than 0.3, it indicates that the adjusted curve maintains good consistency with the typical usage pattern of similar buildings in terms of overall shape. This set of timetable parameters meets the requirements in both mathematical convergence and building functional rationality, and can be used to re-execute the annual hourly energy consumption simulation. When the regularized path distance is greater than or equal to 0.3, it indicates that the shape of the adjusted timetable curve deviates too far from the typical usage pattern. Although this set of parameters may mathematically meet the dual-engine tolerance requirements, the building usage scenario it describes is unreasonable. The self-calibration program will discard this set of parameters and continue to generate the next set of parameters for trial calculation.

[0069] During the self-calibration process, shape constraint checks can be performed on the personnel occupancy rate timetable, lighting utilization rate timetable, and equipment utilization rate timetable separately. Only when the normalized path distance of the curves of the three timetables is less than 0.3 will the set of timetable parameters be used to re-execute the simulation calculation. After traversing all parameter combinations, if no set simultaneously meets the tolerance requirements and shape constraint requirements, the shape deviation tolerance threshold is relaxed first, for example, from 0.3 to 0.4, and the generated parameter combinations are re-screened until a parameter combination that meets the conditions is found.

[0070] In common energy consumption simulation self-calibration methods based on dual-engine cross-validation, the self-calibration process typically aims solely at minimizing the output deviation between the two simulation engines, automatically adjusting input parameters using mathematical optimization algorithms without constraining the rationality of the adjusted parameters in relation to building function. This purely mathematically driven self-calibration may lead to significant deviations of the adjusted timetable parameters from the actual building usage patterns. Although the output deviation between the two engines is eliminated, the calibrated model cannot accurately reflect the actual operating conditions of the building. The proposed solution introduces a shape constraint mechanism using a typical baseline curve as a reference during the calibration process. It utilizes a dynamic time warping algorithm to quantify the degree of shape deviation between the adjusted timetable curve and the typical curve, and uses a preset shape deviation tolerance threshold as a filtering condition. This ensures that the self-calibration results maintain consistency between the timetable parameters and the building's functional usage while satisfying the mathematical convergence of the dual engines.

[0071] According to another embodiment of the present invention, in the provided BIM-based green building design method, when calculating the normalized path distance between the adjusted timetable curve and the corresponding typical benchmark curve using a dynamic time warping algorithm, and comparing the normalized path distance with a preset shape deviation tolerance threshold, the daily usage patterns of different building types are fundamentally different. A single fixed shape deviation tolerance threshold is difficult to adapt to the varying requirements of different building types for timetable flexibility. Office buildings have strong regularity in daily use, with personnel time concentrated during weekday working hours. The fluctuation range of the timetable curve shape is relatively small, and a stricter shape deviation tolerance threshold should be adopted. Commercial buildings have relatively dispersed personnel flow patterns. The personnel attendance rate curve during business hours is flatter than that of office buildings, and the difference between weekends and weekdays is smaller. The timetable curve shape has greater flexibility, and a relatively lenient threshold should be adopted. The usage patterns of educational buildings are affected by course schedules and winter and summer vacations. The timetable curve has obvious semester periodicity and holiday trough characteristics. Its shape flexibility is between that of office buildings and commercial buildings or even more flexible, and a more lenient threshold should be adopted. To accommodate the above differences, the shape deviation tolerance threshold is set as a tiered threshold, and a threshold tiering call step is performed before comparing the regularized path distance with the shape deviation tolerance threshold.

[0072] The threshold grading process retrieves the shape deviation tolerance level corresponding to the building type obtained from the IfcProject attribute by querying a pre-defined threshold grading mapping table. The building type field in the IfcProject attribute can be filled in by the designer during BIM model creation or written by the aforementioned semantic matching model during the building type inference phase. The pre-defined threshold grading mapping table stores the mapping relationship between three common public building types and three shape deviation tolerance levels: office buildings correspond to the first threshold, commercial buildings to the second threshold, and educational buildings to the third threshold, with the first threshold being less than both the second and third thresholds. Specific values ​​can be set as follows: the first threshold (for office buildings) is 0.25, the second threshold (for commercial buildings) is 0.40, and the third threshold (for educational buildings) is 0.55. These values ​​can be determined based on the statistical variance of typical timeline curves for various building types in ASHRAE Standard 90.1, or calibrated using engineering experience data from the field of building energy consumption simulation.

[0073] When executing the timetable shape constraint step, the building type string is first read from the IfcProject attribute, and then a matching item is searched in the graded threshold mapping table. If a matching item is found, such as the building type being "office building", the first threshold of 0.25 is used as the shape deviation tolerance threshold for this self-calibration process. If the building type in the IfcProject attribute is empty, or if the building type string has no matching item in the graded threshold mapping table, the second threshold of 0.40 can be used as the default shape deviation tolerance threshold to ensure that the self-calibration process can still be executed normally even if the building type information is missing. The graded threshold is then passed to the shape constraint verification stage and compared with the warped path distance calculated by the dynamic time warping algorithm. Only when the warped path distance is less than the threshold of the corresponding level is the set of timetable parameters adopted and used to re-execute the annual hourly energy consumption simulation.

[0074] In common energy consumption simulation self-calibration methods, the shape constraints for adjusting timetable parameters typically use fixed thresholds, failing to differentiate the varying needs of different building types for flexible usage patterns. If the fixed threshold is set too strictly, timetable self-calibration for commercial and educational buildings may struggle to find suitable parameter combinations due to the overly tight threshold; conversely, if the fixed threshold is set too loosely, calibration results for office buildings may deviate significantly from actual usage patterns. The proposed solution automatically applies different levels of shape deviation tolerance thresholds based on building type, ensuring that the stringency of shape constraints matches the regularity of actual building usage patterns, thus balancing the convergence feasibility of self-calibration with the functional rationality of the calibration results.

[0075] According to another embodiment of the present invention, in the provided BIM-based green building design method, after the grading threshold is applied according to the building type, differences in personnel activity patterns still exist between different floors and different orientation areas within the same building. For example, the ground floor of a mixed-use office building typically houses a lobby, commercial ancillary spaces, or external service windows. The personnel flow patterns in these areas are significantly influenced by external foot traffic, and their schedule flexibility is significantly higher than that of standard office floors. Conversely, the usage patterns of administrative or executive office areas on the top floor are often more regular than those of standard floors, and their schedule flexibility is lower than that of standard office floors. If the grading threshold corresponding to the building type is uniformly applied to all spaces in the entire building, it may not accurately reflect the differences in usage patterns between different areas within the same building. To refine this difference, a fine-tuning step based on spatial area grading is performed after the threshold grading application step.

[0076] The floor identifier and window orientation identifier of each IfcSpace are extracted according to the spatial region-based fine-tuning steps. The floor identifier can be obtained from the entity attributes of the IfcBuildingStorey where the IfcSpace is located, or it can be calculated from the elevation parameters of the IfcSpace. The window orientation identifier can be extracted from the geometric information of the IfcWindow component associated with that IfcSpace. By calculating the angle between the outer normal direction of the window component and the due north direction, the orientation is classified into one of the four basic directions: east, south, west, and north. For IfcSpaces without windows, their window orientation identifier can be marked as null; such spaces do not participate in the regional fine-tuning based on orientation.

[0077] The fine-tuning rules are preset as follows: for IfcSpaces on the ground floor with street-facing facades, the shape deviation tolerance threshold is increased by one level; for IfcSpaces on the top floor, the shape deviation tolerance threshold is decreased by one level. The ground floor is determined by the floor number corresponding to the floor identifier being equal to 1. The top floor is determined by the floor number corresponding to the floor identifier being equal to the total number of IfcBuildingStoreys. The street-facing facade can be determined based on the relationship between the IfcSpace's window orientation and the building's boundary line or site setback line, or by querying the list of street-facing facade identifiers entered by the designer. When the facade corresponding to the window orientation is marked as a street-facing facade, the IfcSpace is considered a street-facing facade space. The specific meaning of raising a level is as follows: If the space was assigned the first threshold (0.25) in the tiered threshold call step, then after raising a level, the second threshold (0.40) will be used; if it was originally the second threshold (0.40), then after raising it, the third threshold (0.55) will be used; if it was originally the third threshold (0.55), then it will remain at the third threshold without further raising. The same logic applies to lowering a level: if it was originally the second threshold, it will be lowered to the first threshold; if it was originally the third threshold, it will be lowered to the second threshold; if it was originally the first threshold, it will remain at the first threshold without further lowering. When an IfcSpace simultaneously meets both the conditions of being on the ground floor facing the street and on the top floor (e.g., a building with only one floor and a top-floor office space facing the street), the upward adjustment operation will be prioritized because the elastic influence of pedestrian flow on the ground floor is usually stronger than the regular constraints of the top floor.

[0078] The adjusted grading thresholds replace the original shape deviation tolerance thresholds obtained by the corresponding IfcSpace in the threshold grading call step, and are used in subsequent comparisons with the regularized path distance. For IfcSpaces that do not meet the conditions of being on the ground floor facing the street and not meeting the conditions of being on the top floor, the original grading thresholds remain unchanged. After regional fine-tuning, IfcSpaces on different floors within the same building may use different levels of shape deviation tolerance thresholds. The ground floor facing the street uses a more lenient threshold to accommodate the higher flexibility of pedestrian activity in the area, while the top floor uses a stricter threshold to match its more regular usage patterns.

[0079] Existing methods for setting shape deviation tolerance thresholds based on building type typically apply a uniform threshold to all spaces within the same building, failing to consider internal differences in activity patterns across different floors and orientations. The proposed solution, building upon the building type-based threshold, further fine-tunes the threshold by adjusting it upwards or downwards based on the space's floor and orientation. This ensures that the strictness of shape constraints across different functional areas within the same building matches the regularity of their actual usage patterns, maintaining self-calibration convergence while improving the adaptability of calibration results to different regional conditions.

[0080] Application examples A twelve-story office building in Beijing, a cold climate zone, has a total floor area of ​​approximately 18,000 square meters. Its main functions are administrative offices and meeting rooms, and it faces due south. A relatively detailed BIM model was created during the project design phase. The model, exported in IFC format, includes geometric information of building components, thermal performance parameters of the building envelope, and IfcSpace objects for each space. Specifically, the exterior walls consist of 200mm thick aerated concrete blocks covered with 50mm thick extruded polystyrene boards, with a heat transfer coefficient of approximately 0.38 W / m² Kelvin. The windows are double-glazed, low-E (low-emissivity) aluminum alloy windows, with a heat transfer coefficient of approximately 1.8 W / m² Kelvin and a solar heat gain coefficient of approximately 0.45. The roof has a heat transfer coefficient of approximately 0.30 W / m² Kelvin. The IFC file also stores the names of each IfcSpace, such as "Standard Office A," "Open Office Area," "Small Meeting Room," and "Corridor."

[0081] During the energy consumption simulation preparation phase, the program extracts geometric information from the IFC file and converts it into geometric description data in gbXML format. During extraction, it was discovered that some interior partitions lacked thickness information. The program retrieved this information from a pre-set building component information database based on component type and spatial location, supplementing the interior partition thickness to 100 mm. When extracting the thermal performance parameters of the building envelope, it was detected that several exterior windows lacked solar heat gain coefficients. The program searched for these parameters in a pre-set building material thermal parameter database based on the window glass type. For Low-E double-glazed windows, the default solar heat gain coefficient was set to 0.45 and entered into the corresponding field in the gbXML file. Subsequently, the program extracted the name and description attributes of each IfcSpace and used a semantic matching model trained on a building space corpus to determine the functional type of each space. For example, "Standard Office A" and "Open Office Area" both matched as "Office," "Small Meeting Room" matched as "Meeting Room," and "Corridor" matched as "Corridor." The program also read the building type from the IfcProject attributes, which was already filled in as "Office Building" in the IFC file. Using "office building" and "office" as indexes, the personnel density of office space is obtained from the internal thermal disturbance parameter rule base as 0.1 people per square meter, lighting power density as 9 watts per square meter, and equipment power density as 15 watts per square meter. The office personnel attendance rate is as follows: 0:00-7:00 0%, 7:00-8:00 20%, 8:00-12:00 90%, 12:00-13:00 30%, 13:00-17:00 90%, 17:00-19:30%, 19:00-24:00 0% on rest days. Lighting and equipment usage rates are set based on the personnel attendance rate, with the peak equipment usage rate set at 70%. These parameters, along with the corresponding Schedule objects, are written into the Space elements of the gbXML format data.

[0082] After completing the parameter completion, the gbXML file was imported into the main energy consumption simulation engine, EnergyPlus, and the auxiliary energy consumption simulation engine built based on the Modelica Buildings Library. Both engines used the same hourly meteorological data for a typical meteorological year in Beijing and independently performed hourly energy consumption simulations throughout the year. After the simulation, EnergyPlus output the first energy consumption index data, i.e., annual heating and cooling energy consumption per unit building area of ​​48.5 kWh per square meter, while the Modelica model output the second energy consumption index data of 42.1 kWh per square meter. The normalized mean bias between the two was calculated to be 14.8%, and the root mean square error coefficient of variation was 18.6%. The preset engineering tolerance range is a normalized mean bias absolute value of less than 15% and a root mean square error coefficient of variation of less than 20%. The results of this calculation show that both indicators are within the tolerance range; therefore, the simulation results are deemed reliable, and the first energy consumption index data of 48.5 kWh per square meter is taken as the final energy consumption index data.

[0083] If the normalized mean bias of the current calculation result is 16.5%, exceeding the tolerance range, model self-calibration is triggered. The program extracts the timetables for occupancy rate, lighting utilization rate, and equipment utilization rate from the two engines and adjusts them synchronously in 5% increments. In the current model, the peak occupancy rate is 90%, the peak lighting utilization rate is 80%, and the peak equipment utilization rate is 70%. The adjustment ranges are set to 70% to 100% for peak occupancy rate, 60% to 100% for peak lighting utilization rate, and 50% to 90% for peak equipment utilization rate, respectively. For each set of timetable parameter combinations generated, the program first uses a dynamic time warping algorithm to calculate the warped path distance between the adjusted timetable curve and the typical baseline curve of the office building. The shape deviation tolerance threshold for the office building is set to the first threshold of 0.25 according to the graded threshold mapping table. At the same time, for the ground floor street-facing space, the program raises its threshold by one level to 0.40 according to the regional fine-tuning rules, and lowers it by one level to 0.25 for the top floor space (if the top floor was originally at the first threshold, it remains at 0.25). Only when the regularized path distances of the three timeline curves are all less than the shape deviation tolerance thresholds for their respective spaces are these parameters used to re-execute the dual-engine energy consumption simulation. After a thorough iteration, a set of timeline parameters that meet both tolerance requirements and shape constraints is found: peak occupancy rate is adjusted to 88%, peak lighting utilization rate to 82%, and peak equipment utilization rate to 68%. Using this as input, the simulation is repeated, and the dual-engine deviation falls within the tolerance range. The first energy consumption index data of 48.2 kWh per square meter at this point is used as the final energy consumption index data.

[0084] Taking the final energy consumption index of 48.2 kWh per square meter as an example, this value exceeds the preset upper limit of 45 kWh per square meter for green building energy consumption in cold regions. The program decomposes the excess of 3.2 kWh per square meter to each building envelope component. Based on the area, heat transfer coefficient, and temperature difference-driven load contribution rate of each component, the south-facing windows contribute the most, reaching 32%, followed by the roof at 18%, and the south-facing exterior walls at 12%. The program filters components with a contribution rate exceeding 10% to generate a list of key components, including south-facing windows, roof, and south-facing exterior walls. According to the preset high-performance component parameter library, the current heat transfer coefficient of the south-facing windows (1.8 W / m² Kelvin) corresponds to a better first-level parameter of 1.5 W / m² Kelvin, the current heat transfer coefficient of the roof (0.30 W / m² Kelvin) corresponds to a better first-level parameter of 0.25 W / m² Kelvin, and the heat transfer coefficient of the south-facing exterior walls (0.38 W / m² Kelvin) corresponds to a better first-level parameter of 0.30 W / m² Kelvin. The program generates a structured modification instruction containing a unique identifier for the component, the name of the parameter to be modified, the current value, and the target value. It then writes the target parameter value into the corresponding component attribute field of the BIM model through the application programming interface of the BIM modeling software, and then automatically triggers a new round of gbXML export and energy consumption simulation.

[0085] If, in subsequent iterations, it is detected that the heat transfer coefficient of the south-facing window has reached the upper limit of 1.0 W / m² Kelvin for the same type of component in the high-performance component parameter library, but the final energy consumption index is still 46.0 kWh / m², exceeding the threshold of 1.0 kWh / m², a passive adjustment mechanism is triggered. The program locks the thermal parameters of the south-facing window at 1.0 W / m² Kelvin, extracts the building orientation corresponding to the window component as due south, and identifies the facade as the south facade. Dividing the excess energy consumption difference of 1.0 kWh / m² by the comprehensive energy efficiency ratio of the air conditioning system (3.0), the target value for the equivalent solar radiation heat gain reduction is approximately 0.33 kWh / m² / year. Based on the mapping relationship of south-facing orientation in cold regions in the passive component parameter adjustment rule library, this reduction corresponds to a window-to-wall ratio reduction of 0.02 to 0.05, or the addition of horizontally adjustable external shading louvers with a cantilever length of 0.3 to 0.6 meters. The program prepares to generate a window-to-wall ratio modification instruction, but performs a natural lighting compliance check before generation. The geometric boundaries of the standard office to which the window belongs were extracted from the BIM model. The room's net dimensions are 6m × 8m × 3.6m, the reference work surface height is 0.75m, and the visible light transmittance is 0.62. Based on Beijing meteorological data, the percentage of the room's total natural lighting time is calculated to be 55%, which is lower than the preset green building lighting standard threshold of 60%. Therefore, the program rejects the window-to-wall ratio modification instruction, retains the original window-to-wall ratio, and prioritizes the use of adjustable external shading component size parameters, initially selecting adjustable external shading louvers with a length of 0.5m.

[0086] When generating the new instruction for the external shading component, the cross-space shading coordination step detected that the adjacent meeting room window (2.1 meters wide and 2.1 meters high) and the current office window (3.0 meters wide and 2.1 meters high) on the south facade are located on the same facade, with a center distance of 2.8 meters. The adjacent meeting room window is projected parallel to the current office window, and the calculated overlap area is approximately 22%, which is less than the shading threshold of 25%. Therefore, no reduction is performed, and the new instruction for the external shading component is generated directly with a cantilever length of 0.5 meters. If three adjacent spaces on the south facade simultaneously generate new instructions for external shading components, with window areas of 6.3 square meters for the office, 4.4 square meters for the meeting room, and 3.8 square meters for the administration office, the program initiates the adjacent space coordination master control, prioritizing the office with the largest window area and executing its external shading component instruction first, using its 0.5-meter cantilever length as the baseline shading parameter for that facade. When generating their respective instructions, the meeting room and administrative office used the baseline occlusion parameter of 0.5 meters and the geometric position of the office windows as known boundary conditions to calculate and reduce the overlapping area. Ultimately, the cantilever length of the meeting room was reduced to 0.45 meters, and that of the administrative office to 0.40 meters. After all instructions were written into the BIM model, the simulation was triggered again until the energy consumption targets were met, completing the entire design optimization loop.

[0087] The number of devices and processing scale described herein are for the purpose of simplifying the description of the invention. Applications, modifications, and variations of the invention will be readily apparent to those skilled in the art.

[0088] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. They can be applied to various fields suitable for the present invention. For those skilled in the art, other modifications can be easily made. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details and embodiments shown and described herein.

Claims

1. A BIM-based green building design method, characterized in that, Includes the following steps: Step 1: Construct the BIM model of the target building. The BIM model is an IFC format file and contains geometric information of building components, thermal performance parameters of the building envelope, building material property information, and spatial objects expressed by IfcSpace instances. Step 2: Extract geometric information and thermal performance parameters of the building envelope from the IFC format file, convert the geometric information into gbXML geometric description data, map the thermal performance parameters of the building envelope to gbXML material attribute fields, and complete any missing geometric information and thermal performance parameters. Step 3: Extract the name and description attributes of IfcSpace instances from the IFC format file, and use a pre-trained semantic matching model to determine the spatial function type corresponding to each IfcSpace; obtain the building type from the IfcProject attribute. When this attribute is empty, determine the building type based on the area ratio of each spatial function type; using the building type and spatial function type as indexes, obtain the personnel density, lighting power density, equipment power density, and corresponding schedule for each IfcSpace from the preset internal thermal disturbance parameter rule base, and write them into gbXML format data. Step 4: Import the gbXML format data after completing the geometric information, thermal performance parameters and internal thermal disturbance parameters into the main energy consumption simulation engine and the auxiliary energy consumption simulation engine based on different mathematical mechanisms, respectively, and independently execute the hourly energy consumption simulation throughout the year, and output the first energy consumption index data and the second energy consumption index data respectively. Step 5: Calculate the normalized mean bias and root mean square error coefficient of variation between the first energy consumption index data and the second energy consumption index data. When both the normalized mean bias and the root mean square error coefficient of variation are within their respective preset engineering tolerance thresholds, the first energy consumption index data is used as the final energy consumption index data; otherwise, the model self-calibration is performed until the tolerance requirements are met. Step Six: Compare the final energy consumption index data with the preset green building energy consumption threshold range. When the final energy consumption index data exceeds the green building energy consumption threshold range, decompose the excess energy consumption to each building envelope component to form a key component list. Generate component-level structured modification instructions based on the preset high-performance component parameter library and write them into the BIM model through the application programming interface of the BIM modeling software. This will automatically trigger a new round of gbXML file export and building energy consumption simulation calculation.

2. The BIM-based green building design method as described in claim 1, characterized in that, When it is detected that the current thermal performance parameter of a component in the critical component list has reached the upper limit of the performance of the same type of component in the preset high-performance component parameter library, and the final energy consumption index data still exceeds the green building energy consumption threshold range, the following steps are further executed: The parameters of the component are locked as the upper limit of performance. The building orientation and the facade identification of the component are extracted. The excess energy consumption difference of the component that exceeds the green building energy consumption threshold range in the energy consumption excess decomposition in step six is ​​converted into an equivalent solar radiation heat gain reduction target value. The equivalent solar radiation heat gain reduction target value is obtained by dividing the excess energy consumption difference by the preset comprehensive energy efficiency ratio of the air conditioning system. Based on the target value of solar radiation heat gain reduction, the corresponding window-to-wall ratio adjustment range or external shading component size parameters are queried from the preset passive component parameter adjustment rule library, and a window-to-wall ratio modification instruction or external shading component addition instruction containing a unique identifier of the window component and an exterior facade identifier is generated. The window-to-wall ratio modification command or the external shading component addition command is written into the BIM model through the application programming interface of the BIM modeling software, triggering a new round of building energy consumption simulation calculation.

3. The BIM-based green building design method as described in claim 2, characterized in that, Before generating the window-to-wall ratio modification command, a natural lighting compliance verification step is also included: Extract the geometric boundary information and visible light transmittance parameters of the space to which the window component belongs from the BIM model. Based on the sky illuminance data in the typical meteorological year file, calculate the percentage of full natural lighting time on the reference working plane of the space. When the percentage of total natural daylighting time is lower than the preset threshold for green building daylighting compliance, the window-to-wall ratio modification instruction is rejected, the original window-to-wall ratio parameter is retained, and the adjustable external shading component size parameter corresponding to the window component is preferentially retrieved from the passive component parameter adjustment rule library according to the target value of solar radiation heat gain reduction. The external shading component addition instruction is regenerated and written into the BIM model.

4. The BIM-based green building design method as described in claim 3, characterized in that, When retrieving and calling the adjustable external shading component size parameters corresponding to the window component from the passive component parameter adjustment rule base, a cross-space shading coordination step is also included: Extract the location coordinates and geometric dimensions of the window component in the adjacent space on the same facade as the window component from the BIM model; Project the window components of the adjacent space into the current verification space in parallel, and calculate the percentage of the overlapping area between the projected area and the window components of the current verification space. When the overlap area exceeds the preset shading threshold, the size parameters of the retrieved adjustable external shading component are reduced according to the preset reduction coefficient, and a new instruction for external shading component is generated based on the reduced size parameters.

5. The BIM-based green building design method as described in claim 4, characterized in that, When multiple adjacent spaces on the same facade simultaneously generate new instructions for external shading components, the process also includes a main control step for coordinating adjacent spaces: Extract the window area parameters of all external shading components to be added on the facade from the BIM model, and determine the space with the largest window area as the dominant space of the facade. Prioritize the execution of new external shading components for the dominant space, and use the size parameters of the external shading components for the dominant space as the reference shading parameters for that facade. When generating new instructions for external shading components in non-dominant spaces, the reference shading parameters are substituted as known boundary conditions into the cross-space shading coordination step for reduction calculation, and the reduced size parameters are used to generate new instructions for external shading components in non-dominant spaces.

6. The BIM-based green building design method as described in claim 1, characterized in that, Step five, model self-calibration, includes: Extract the time parameters used to define the occupancy rate, lighting utilization rate, and equipment utilization rate from the main energy consumption simulation engine and the auxiliary energy consumption simulation engine, respectively. Within the percentage range of their respective timetable parameters, the timetable parameters in the two simulation engines are adjusted synchronously with a preset step size to generate multiple sets of timetable parameter combinations and re-execute the annual hourly energy consumption simulation until the calculated normalized mean error and root mean square error coefficient of variation are both within their respective preset engineering tolerance threshold ranges. The final set of timetable parameters that meets the tolerance requirements, along with their corresponding first energy consumption index data, will be used as the final reliable simulation input and output results.

7. The BIM-based green building design method as described in claim 6, characterized in that, When synchronously adjusting the timetable parameters in the two simulation engines with preset step sizes, a timetable shape constraint step is also included: Obtain the typical occupancy rate baseline curve, typical lighting utilization rate baseline curve, and typical equipment utilization rate baseline curve corresponding to the building type; After each timetable parameter combination is generated, the dynamic time warping algorithm is used to calculate the warped path distance between the adjusted timetable curve and the corresponding typical baseline curve. The regularized path distance is compared with a preset shape deviation tolerance threshold. Only when the regularized path distance is less than the shape deviation tolerance threshold is the set of timetable parameters used to re-execute the annual hourly energy consumption simulation.

8. The BIM-based green building design method as described in claim 7, characterized in that, The shape deviation tolerance threshold is a graded threshold, and before comparing the dynamic time warping path distance with the shape deviation tolerance threshold, a threshold grading invocation step is also included: Based on the building type obtained from the IfcProject property, query the shape deviation tolerance level corresponding to the building type in the preset grading threshold mapping table; The hierarchical threshold mapping table stores the mapping relationship between office buildings and the first threshold, commercial buildings and the second threshold, and educational buildings and the third threshold, with the first threshold being less than the second threshold and less than the third threshold.

9. The BIM-based green building design method as described in claim 8, characterized in that, Following the threshold-based hierarchical invocation step, a spatial region-based fine-tuning step is also included: Extract the floor identifier and window orientation identifier of each IfcSpace. According to the preset regional fine-tuning rules, increase the shape deviation tolerance threshold by one level for IfcSpaces on the ground floor and facing the street, and decrease the shape deviation tolerance threshold by one level for IfcSpaces on the top floor. Increasing the threshold by one level means adjusting the current threshold to the next lower threshold with a larger value in the hierarchical threshold mapping table, and decreasing the threshold by one level means adjusting the current threshold to the next lower threshold with a smaller value. Replace the original shape deviation tolerance threshold of the corresponding IfcSpace with the adjusted hierarchical threshold for subsequent comparison with the regularized path distance.